Your new hire takes 90 days. What if it took 9?
New hires spend their first three months reconstructing the context the team already has — Slack archaeology, deprecated docs, “ask anything” channels. engr4m turns onboarding into a graph walk. Cited, current, self-serve from day one.
Day one. Welcome to the cold start.
Your new hire — let's call them Sam — joined Monday. By Tuesday afternoon they have access to Slack, Notion, Jira, GitHub, Linear, Calendar and four shared drives. By Wednesday they've been added to sixteen channels and invited to twenty-two intro 1:1s. By Friday they've asked their first question in #ask-anything — and gotten three different answers from three different people, two of which contradict each other.
Week two: Sam reads Notion. Half the docs are deprecated; the other half were written for an audience that no longer exists. They scroll Slack history for the renewal pipeline, find six channels mentioning it, none with a complete picture. By Friday they've absorbed enough to ask a sharp question — and triggered a forty-minute side conversation between two senior engineers about which spec is current.
Week ten: Sam is shipping. They've finally internalized the team's working knowledge. The cost: ten weeks of low output, one senior engineer's afternoon a week answering basic questions, and a ramp-up tax that nobody on the team can fully account for.
8–12 weeks
Average time for a new individual contributor to reach full productivity in a knowledge-work role. The biggest predictor isn't talent — it's how reachable the team's existing knowledge is.
SHRM — Onboarding Research
Your team's brain lives in six tools and four heads
Onboarding is supposed to be a document. In practice it's archaeology. The decisions are in Slack threads. The status is in Jira. The rationale is in a half-finished Notion page. The actual current plan is in the head of the engineer who's on vacation this week. The new hire's job, for the first three months, is to reconstruct from these fragments the picture the rest of the team already has.
Onboarding isn't a document. It's a graph the new hire can walk through.
And so the cost of every new hire is paid twice: once in their salary during the ramp, and once in the productive time of the people teaching them. The senior engineer who answers the same question for the third time isn't being inefficient — they're being a human cache for knowledge the tools should have surfaced.
Day one with engr4m: hand them the graph
Sam joins on Monday. By Monday afternoon they have access to engr4m. They see every active case scoped to their team. They can ask any question and get an answer cited back to the source. They start their week reading the actual conversations that produced the current plan — not a stale summary, not a doc someone forgot to update.
Three active cases on the team this week. Renewal pipeline has a deadline in 14 days. API migration is mid-rollout. Two open customer escalations need owners.
Active cases
- Acme renewal — JaneCase
- API migration · phase 2Case
- Pricing v3 specCase
This week's wins
- Auth rotation shippedGitHub
- Outlook integration liveSlack
Open questions
- Cutover date for v3Slack
- Owner for SOC2 audit prepJira
Sam picks the renewal pipeline case — the one with the deadline. They want to understand its current state without interrupting Jane. They ask the graph.
The Acme case is linked to four events: a customer email about pricing v3 from Tuesday, a #revenue thread from Wednesday about the v3 spec, ticket DEAL-119 owned by Jane, and last Thursday's renewal sync.
Four citations. Sam clicks each one in turn and reads the source. By the end of the day they understand the renewal pipeline as well as anyone on the team — and they've done it without interrupting a single person. Wednesday morning they message Jane: “I saw the open question in #revenue. Want me to draft the spec response?”
- Sam reconstructs context from six tools for 10 weeks.
- Senior engineers lose an afternoon a week to repeat questions.
- Notion docs decay; the “truth” lives in heads.
- First real contribution at week 8 or later.
- Sam reads the actual conversations on day one.
- Self-serve answers — no senior-engineer-as-cache.
- Provenance preserved; the graph stays current.
- First real contribution within the first week.
Why onboarding through the graph is different
Cases scope the world for them
Every active piece of work is a case. Sam sees what the team is actually doing this week, not a wiki page from 2024.
Ask AI gives cited answers
No hallucinations and no stale summaries — every answer points back to the exact Slack message or PR that produced it.
Provenance, not paraphrase
Sam reads the original conversation. They learn the team's reasoning, not someone's edit of it.
Senior engineers stop being a cache
The same questions get answered automatically by the graph instead of by interrupting your most expensive engineers.
Daily Briefing keeps them current
Sam wakes up to the same brief the leads get — work done, focus next, open questions and blockers. They're never out of the loop.
Contribute on day three, not day ninety
When the graph answers the questions, the new hire spends their first week shipping — not auditioning for productivity.